Hierarkkisen vahvistusoppimisen soveltuvuuden arviointi videopelien kehittämisessä
Bibliographic record
Abstract
In this thesis we explore the feasibility of using hierarchical reinforcement learning (HRL) in video game development to create non-player characters (NPC). NPCs are a crucial part of video games affecting many parts of the game, including storytelling, atmosphere, and importantly work as opponents and teammates. Using traditional methods to create NPCs in video games can be a lengthy and difficult process requiring expert knowledge. Reinforcement learning (RL) has shown potential, but has remained largely unused in video game development due to some major issues. HRL provides solutions to these issues, allowing the complex task to be split into smaller, easier to learn sub-tasks. We design, implement, and study a new HRL method with the potential of creating NPCs with multiple competency levels with minimal effort. Our design is based on a goal-conditional framework which we modify to suit our goals. Instead of using a goal-vector we repurpose it to a skill-vector, which could allow us to mask it and re-train the higher-level policy to prevent certain skills from being used. In order to experiment with our HRL method, we create an physics based quadruped locomotion environment that has possibility for learning multiple different skills. We evaluate our method with and without information hiding in attempt to force certain types of behaviours for the policy levels. The method shows potential in our experiments but requires further experimentation and engineering to create multiple competency levels.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.003 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.004 | 0.002 |
| Open science | 0.001 | 0.003 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.033 | 0.010 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".